I have built ghc-7.4.1 on two servers but it keeps failing on a third with: make[1]: *** [libraries/haskell2010/dist-install/build/Prelude.dyn_o] Segmentation fault make[1]: *** Waiting for unfinished jobs.... make[1]: *** [libraries/haskell98/dist-install/build/Prelude.p_o] Segmentation fault make[1]: *** [libraries/haskell2010/dist-install/build/Prelude.p_o] Segmentation fault make[1]: *** [libraries/haskell98/dist-install/build/Prelude.dyn_o] Segmentation fault make[1]: *** [libraries/haskell2010/dist-install/build/Prelude.o] Segmentation fault make[1]: *** [libraries/haskell98/dist-install/build/Prelude.o] Segmentation fault make[1]: *** [utils/ghctags/dist-install/build/Main.o] Segmentation fault make: *** [all] Error 2 * ERROR: dev-lang/ghc-7.4.1 failed (compile phase):
Created attachment 306853 [details] Emerge info
Created attachment 306855 [details] emerge -pqv
Please attach the entire build log to this bug report.
Created attachment 306883 [details] Build log This was attached when the bug was opened but seems to have been dropped. Ah, it was too large before compressing it...
Does workaround to disable parallel build helps? MAKEOPTS=-j1 emerge -1 =ghc-7.4.1 Thanks for the report.
(In reply to comment #5) > Does workaround to disable parallel build helps? > > MAKEOPTS=-j1 emerge -1 =ghc-7.4.1 > > Thanks for the report. Hi Sergei, As of today's emerge world it built without problem. I will certainly try the above if I have a problem again (and usually do before posting.) Sorry for the noise. Todd
(In reply to comment #6) > (In reply to comment #5) > > Does workaround to disable parallel build helps? > > > > MAKEOPTS=-j1 emerge -1 =ghc-7.4.1 > > > > Thanks for the report. > > Hi Sergei, > > As of today's emerge world it built without problem. I will certainly try > the above if I have a problem again (and usually do before posting.) > > Sorry for the noise. It's definetly not a noise. Don't be upset! You've caught real bug. Two persons reported almost the same failure. (I don't understand mechanics of the bug though though). It's just very hard to reproduce (the more parallel the build is - the more probable SIGSEGV looks to be).
Created attachment 308747 [details] compressed ghc build log
Created attachment 308749 [details] Another compressed ghc build log I'm getting two variations of the failed build, this and the previous log. It looks like there is some race in parallel build since it failed 1 of 3 times when not using tmpfs for building, but with tmpfs it fails every time in one of the ways.
With MAKEOPTS=-j1 it builds.. but takes a loong time. Portage 2.1.10.56 (default/linux/amd64/10.0/desktop/gnome, gcc-4.6.2, glibc-2.14.1-r2, 3.3.1-gentoo x86_64) ================================================================= System Settings ================================================================= System uname: Linux-3.3.1-gentoo-x86_64-Intel-R-_Core-TM-_i7-2630QM_CPU_@_2.00GHz-with-gentoo-2.1 Timestamp of tree: Fri, 13 Apr 2012 04:45:01 +0000 app-shells/bash: 4.2_p24 dev-java/java-config: 2.1.11-r3 dev-lang/python: 2.6.7-r2, 2.7.3, 3.1.4-r4, 3.2.2-r1 dev-util/cmake: 2.8.7-r5 dev-util/pkgconfig: 0.26 sys-apps/baselayout: 2.1 sys-apps/openrc: 0.9.9.3 sys-apps/sandbox: 2.5 sys-devel/autoconf: 2.13, 2.68 sys-devel/automake: 1.9.6-r3, 1.10.3, 1.11.4 sys-devel/binutils: 2.22-r1 sys-devel/gcc: 4.4.7, 4.5.3-r2, 4.6.2 sys-devel/gcc-config: 1.6 sys-devel/libtool: 2.4.2 sys-devel/make: 3.82-r3 sys-kernel/linux-headers: 3.3 (virtual/os-headers) sys-libs/glibc: 2.14.1-r2 Repositories: gentoo x11 science gamerlay-stable bumblebee local ACCEPT_KEYWORDS="amd64 ~amd64" ACCEPT_LICENSE="* -@EULA" CBUILD="x86_64-pc-linux-gnu" CFLAGS="-march=native -mtune=native -O3 -pipe -ggdb" CHOST="x86_64-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/share/gnupg/qualified.txt /var/lib/hsqldb" CONFIG_PROTECT_MASK="/etc/ca-certificates.conf /etc/dconf /etc/env.d /etc/env.d/java/ /etc/fonts/fonts.conf /etc/gconf /etc/gentoo-release /etc/revdep-rebuild /etc/sandbox.d /etc/splash /etc/terminfo /etc/texmf/language.dat.d /etc/texmf/language.def.d /etc/texmf/updmap.d /etc/texmf/web2c" CXXFLAGS="-march=native -mtune=native -O3 -pipe -ggdb" DISTDIR="/usr/portage/distfiles" EMERGE_DEFAULT_OPTS="" FEATURES="assume-digests binpkg-logs compress-build-logs distlocks ebuild-locks fixlafiles news parallel-fetch parallel-install protect-owned sandbox sfperms splitdebug strict unknown-features-warn unmerge-logs unmerge-orphans userfetch userpriv usersandbox usersync" FFLAGS="-march=native -mtune=native -O3 -pipe -ggdb" GENTOO_MIRRORS="ftp://trumpetti.atm.tut.fi/gentoo/ http://trumpetti.atm.tut.fi/gentoo/ http://gentoo.tups.lv/source/ " LANG="lv_LV.UTF-8" LC_ALL="lv_LV.UTF-8" LDFLAGS="-Wl,-O1 -Wl,--as-needed" LINGUAS="lv en" MAKEOPTS="-j9" PKGDIR="/usr/portage/packages" PORTAGE_CONFIGROOT="/" PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --stats --human-readable --timeout=180 --exclude=/distfiles --exclude=/local --exclude=/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/var/lib/layman/x11 /var/lib/layman/science /var/lib/layman/gamerlay /var/lib/layman/bumblebee /usr/local/portage" SYNC="rsync://rsync.europe.gentoo.org/gentoo-portage" USE="X a52 aac acl acpi alsa amd64 avx bash-completion berkdb bluetooth branding bzip2 cairo cdda cdio cdr cjk cleartype cli colord consolekit cracklib crypt cups cxx dbus dirac djvu dri dts dvd dvdr eds emboss encode evo exif fam ffmpeg fftw firefox flac fontconfig fortran gdbm gdu gif gnome gnome-keyring gnome-online-accounts gphoto2 gpm gsm gstreamer gtk gtk3 iconv idn ipv6 jpeg kate lcms ldap libcaca libnotify live mad matroska mmx mng modules mp3 mp4 mpeg mtp mudflap multilib musepack nautilus ncurses networkmanager nls nptl nptlonly ogg openexr opengl openmp pam pango pcre pdf png policykit ppds pppd pulseaudio qt3support qt4 raw readline schroedinger sdl session smp socialweb speex spell sse sse2 sse4_1 ssl ssse3 startup-notification svg sysfs system-sqlite tcpd theora tiff truetype udev unicode usb v4l v4l2 vaapi vorbis vpx wmf x264 xcb xetex xml xmp xorg xpm xulrunner xv xvid xvmc zlib" ALSA_CARDS="ali5451 als4000 atiixp atiixp-modem bt87x ca0106 cmipci emu10k1x ens1370 ens1371 es1938 es1968 fm801 hda-intel intel8x0 intel8x0m maestro3 trident usb-audio via82xx via82xx-modem ymfpci" ALSA_PCM_PLUGINS="adpcm alaw asym copy dmix dshare dsnoop empty extplug file hooks iec958 ioplug ladspa lfloat linear meter mmap_emul mulaw multi null plug rate route share shm softvol" APACHE2_MODULES="actions alias auth_basic authn_alias authn_anon authn_dbm authn_default authn_file authz_dbm authz_default authz_groupfile authz_host authz_owner authz_user autoindex cache cgi cgid dav dav_fs dav_lock deflate dir disk_cache env expires ext_filter file_cache filter headers include info log_config logio mem_cache mime mime_magic negotiation rewrite setenvif speling status unique_id userdir usertrack vhost_alias" CALLIGRA_FEATURES="kexi words flow plan sheets stage tables krita karbon braindump" CAMERAS="ptp2" COLLECTD_PLUGINS="df interface irq load memory rrdtool swap syslog" ELIBC="glibc" FOO2ZJS_DEVICES="hp1018" GPSD_PROTOCOLS="ashtech aivdm earthmate evermore fv18 garmin garmintxt gpsclock itrax mtk3301 nmea ntrip navcom oceanserver oldstyle oncore rtcm104v2 rtcm104v3 sirf superstar2 timing tsip tripmate tnt ubx" INPUT_DEVICES="evdev synaptics" KERNEL="linux" LCD_DEVICES="bayrad cfontz cfontz633 glk hd44780 lb216 lcdm001 mtxorb ncurses text" LINGUAS="lv en" PHP_TARGETS="php5-3" RUBY_TARGETS="ruby18" USERLAND="GNU" VIDEO_CARDS="dummy fbdev nvidia i965 intel vesa" XTABLES_ADDONS="quota2 psd pknock lscan length2 ipv4options ipset ipp2p iface geoip fuzzy condition tee tarpit sysrq steal rawnat logmark ipmark dhcpmac delude chaos account" Unset: CPPFLAGS, CTARGET, INSTALL_MASK, PORTAGE_BUNZIP2_COMMAND, PORTAGE_COMPRESS, PORTAGE_COMPRESS_FLAGS, PORTAGE_RSYNC_EXTRA_OPTS, USE_PYTHON ================================================================= Package Settings ================================================================= dev-lang/ghc-7.4.1 was built with the following: USE="ghcbootstrap (multilib) -binary -doc -llvm" CFLAGS="-march=native -mtune=native -pipe -ggdb -O2 -march=native -march=native -mtune=native -mtune=native -Wa,--noexecstack -Wa,--noexecstack" CXXFLAGS="-march=native -mtune=native -pipe -ggdb -O2"
Pushed workaround as: > 13 Apr 2012; Sergei Trofimovich <slyfox@gentoo.org> ghc-6.10.4-r1.ebuild, > ghc-6.12.3-r2.ebuild, ghc-6.12.3.ebuild, ghc-7.4.1.ebuild: > Disable parallel make due to build system failures: bug #409631 by Anton > Kochkov, bug #409873 by Todd Goodman. It once again disables parallel building, which is painful, but more reliable.
It is not fixed tough.
(In reply to comment #12) > It is not fixed tough. You mean you have that bit in dev-lang/ghc/ChangeLog and it still fails for you?
No, I mean that setting -j1 is not a fix for parallel build issues. Curiously, I recompiled ghc with -ghcbootstrap and then again with it enabled and now it seems to compile fine with -j9 even on tempfs (3 times in a row).
(In reply to comment #14) > No, I mean that setting -j1 is not a fix for parallel build issues. > > Curiously, I recompiled ghc with -ghcbootstrap and then again with it > enabled and now it seems to compile fine with -j9 even on tempfs (3 times in > a row). That's why it's called a workaround papering over a build failure. The real problem is racing of a bunch of things: - multiple ghc stages rebuilding the same dependencies [unlinely] - multiple ghc jobs building the same library with different flavours [likely] - just bugs in package interdependencies [likely] - bugs in GNU make build system [unlikely] I'm not against the real fix, but the issue is known to upstream since forever. Things keep being fixed, but keep biting users with 5+ fast cores. I was never able to reproduce the issue on 32-headed slow sparc machine, neither it seems to break on 8 headed ppc, so I have nothing to debug :[ Gentoo and upstream will be happy to have more info about build failures and some analysis by users interested in digging the problem up.
If the issue is an old one, then its understandable. It's just nice to have it compiled 3x faster if possible, but probably nicer is if it compiles with the first try :) Well I actually have a 4 core CPU with multithreading, so it is seen as an 8 core CPU, but half of the 'cores' are much slower so I'm overcommiting build jobs a bit. It seems that when some of them get sheduled on the slow cores, they start to lag behind the rest of the builds and it might be that it leads to wrong dependencies or compilation order at some point. It should be possible to test it running with -j4 or 5. The important thing is that upstream is aware off the issue. I'm afraid that there will be little progress on a closed bug in Gentoo bugzilla.
